On 6/1/2018 1:40 PM, Michael Ellerman wrote:=0A=
Scott Wood [off-list ref] writes:=0A=
=0A=
quoted
On Thu, 2018-05-31 at 14:35 +0000, Diana Madalina Craciun wrote:=0A=
quoted
On 5/31/2018 5:21 PM, Michael Ellerman wrote:=0A=
quoted
We can add a nospectre_v1 command line option if necessary.=0A=
What about nobarrier_nospec (or similar) instead of nospectre_v1 comman=
d=0A=
quoted
quoted
line? We are not disabling all the v1 mitigations, the masking part wil=
l=0A=
quoted
quoted
remain unchanged.=0A=
I think nospectre_v1 makes more sense as it's about the user's intention=
s=0A=
quoted
rather than the implementation. The user is giving the kernel permissio=
n to=0A=
quoted
not defend against spectre v1, and it's up to the implementation which=
=0A=
quoted
mitigations (if any) to disable in response to that, same as any other=
=0A=
quoted
optimization.=0A=
Yeah I agree. We also have `nospectre_v2` on x86/s390 so I think keeping=
=0A=
consistency with that is a must.=0A=
=0A=
cheers=0A=
=0A=
OK=0A=
=0A=
Diana=0A=
=0A=